LLMilestonesA milestone timeline of AI and large language models

Act VI · The frontier and its limits

Closed modelOpenAI

GPT-5.5

OpenAI released GPT-5.5 (codename Spud), improving agentic coding and research ability.

82.7%Terminal-Bench 2.0 (reported by OpenAI)

Why it matters

It led agentic-coding tests such as Terminal-Bench 2.0; from May, its Instant version became ChatGPT's default for free users.

Built on

GPT-5

Led to

GPT-6

Source: openai.com ↗

See it on the 3D timeline →